Managing and characterizing complex traits, including quantitative traits

Charlie Brummer Director, Plant Breeding Center The University of California, Davis

UC Davis – Plant Breeding Center

Consolidate and update graduate plant breeding curriculum – include field based content/experience

Involve industry/commodity partners in educational programs through internships, assistantships, class visits, guest lectures….

Promote hiring of more plant breeding faculty and staff with field-based expertise

Develop post-graduate school educational programs, coordinating with Seed Central and Plant Breeding Academy

Develop useful technologies, germplasm and cultivars and actively license for commercialization

Outline

Breeding complex traits

Markers and mapping

Genomic selection

In-field phenotyping

Technology development

High-throughput prediction

108 markers – 10 LG – 1 year

21 years of markers and maps

Have markers contributed to cultivar advances?

Forage cultivars developed using markers

Markers are no longer limiting

Genotyping-by-sequencing to generate saturated map in a month

(Li et al., Manuscript in preparation) 4 haplotypes per chromosome

Bioinformatics-intensive to manipulate sequence data

CC78 SSR markers CC78 GBS SNPs

Chr7 32-38 Mbp

Chr1 Chr2 Chr3 Chr4 Chr5 Chr6 Chr7 Chr8

(Stanton-Geddes et al., 2013)

A flowering time gene on chr7 in M.truncatula MtFD (34.3 Mbp)

Flowering time in alfalfa

Aligned to Medicago whole genome sequence

Fall dormancy in alfalfa

• Robust field-based protocol gives reliable results

• Based on well-defined check cultivars

Taller plants under acclimating conditions are less dormant

Cultivated alfalfa germplasm structured by fall dormancy

Fall Dormancy Breeding Germplasm Population 21 breeding populations 7500 Infinium SNP loci

(Li et al., 2014 PlosOne)

Dormant

Non-dormant

Semi-dormant

R2=64.2%

Chr1 Chr2 Chr3 Chr4 Chr5 Chr6 Chr7 Chr8

Autumn dormancy mapping

DM35

Bi-parental population

FD GWAS population

Chr1 Chr2 Chr3 Chr4 Chr5 Chr6 Chr7 Chr8

Breeding germplasm population

Li & Brummer, 2012, Agronomy 2:40

Predicting phenotypes with markers

Generate genome-wide markers and compute the effect of each marker on the trait

Select based on the sum of these affects across all markers – the Genome Estimated Breeding Value

Measure phenotypes on a plant population

1-A

1-B

2-A

2-B

3-A

3-B

4-A

4-B

5-A

5-B

Year & season

R

E

E

E

S&R

E

E

E

S&R

E

Phenotypic selection

R

E

E

E

E

E

E

E

S&R

S&R

S&R

S&R

S&R

S&R

Build model

Update model

Genomic selection

Alfalfa Genomic Selection

NECS breeding population – Clonal selection

2 cycles, ~200 genotypes/cycle, 3-5 yield environments/cycle

GBS (Elshire protocol) 96-plex on Hi-Seq 2000 ~2 million reads/genotype rrBLUP (Endelman)

Li, Wei, Acharya, Viands, Hansen, Claessens, Brummer – unpub.

Iowa New York Quebec

Cycle 0 0.44 0.49 0.43

Cycle 1 - 0.38 0.62

C0 – C1 - 0.44 0.14

Accuracy (r) within locations

0

0.1

0.2

0.3

0.4

0.5

Iowa New York Quebec

Acc

ura

cy (

r)

Iowa New York Quebec

Location cross-validation – Cycle 0

Predict yield based on marker breeding values

Markers for disease/insect resistance

Effective greenhouse (phenotypic) screen

Marker confirmation during breeding program

C+7

/C-1

Chromosome 5 Chromosome 6 Chromosome 7 Chromosome 8

Confirmed by SNP markers on individual genotypes (Monteros)

Waiting for phenotypic data….

Tracking allele frequency changes

>55,000 SNP loci with >100 reads/pop; mapped to Medicago

GBS using ~100 bulked genotypes of C-1, C+4 and C+7

Whitefly resistance – Larry Teuber (UCD)

High-throughput phenotyper (Jesse Poland, USDA-ARS, Manhattan, KS)

COLORADO: Canopy spectral reflectance (CSR) measured with a Jaz spectroradiometer

CSR measures the amount of light reflected from the plant canopy at many wavelengths

High Throughput Phenotyping (HTP)

Wheat breeding nursery, Jesse Poland, USDA-ARS, Manhattan, KS, USA

Geo-referenced data collection

Non-destructive measurements

Fast, repeatable

Assuming good calibrations

Possible Uses in Breeding

Predict yield before harvest Measure water, nutrient status Collect data at more locations than possible with typical trait measurements

Wheat breeding nursery, Jesse Poland, USDA-ARS, Manhattan

High Throughput Phenotyping Platforms

UCD participants in HTP

• CAES

– M. Gilbert (crop physiology), J. Ross-Ibarra (genomics), P. Gepts (genomics/breeding), C. Brummer (Plant Breeding Center), A. Walker (grape breeding), R. Hijmans (geospatial analysis)

• CES

– S. Vougioukas (actuators, sensors and control systems), M. Delwiche (biosensors, electr instrum), Computer Sci: (data processing, large datasets, telematics)

• New positions being requested from university

Summary

High-throughput genotyping – available now

High-throughput phenotyping – predict phenotypes

area of interest in many crops

many questions – how well do sensors predict desired traits

Data handling

informatics to manipulate raw sequence or phenotype data

data processing/analysis power

data storage and retrieval